STUDY PROTOCOL Open Access

Lessening Organ dysfunction with VITaminC (LOVIT): protocol for a randomizedcontrolled trialMarie-Hélène Masse1, Julie Ménard1, Sheila Sprague2,3, Marie-Claude Battista4, Deborah J. Cook3,Gordon H. Guyatt3, Daren K. Heyland5, Salmaan Kanji6, Ruxandra Pinto7, Andrew G. Day8, Dian Cohen9,Djillali Annane10, Shay McGuinness11,12, Rachael Parke11,12,13, Anitra Carr14, Yaseen Arabi15,Bharath Kumar Tirupakuzhi Vijayaraghavan16, Frédérick D’Aragon4, Élaine Carbonneau1, David Maslove17,Miranda Hunt17, Bram Rochwerg18, Tina Millen19, Michaël Chassé20, Martine Lebrasseur20, Patrick Archambault21,22,Estel Deblois22, Christine Drouin26,28, François Lellouche23, Patricia Lizotte24, Irene Watpool25, Rebecca Porteous25,France Clarke26, Nicole Marinoff7, Émilie Belley-Côté27, Brigitte Bolduc28, Scott Walker29, John Iazzetta30,Neill K. J. Adhikari7,31*† , François Lamontagne4*† and on behalf of the Canadian Critical Care Trials Group

Abstract

Background: Sepsis is a health problem of global importance; treatments focus on controlling infection andsupporting failing organs. Recent clinical research suggests that intravenous vitamin C may decrease mortality insepsis. We have designed a randomized controlled trial (RCT) to ascertain the effect of vitamin C on the compositeendpoint of death or persistent organ dysfunction at 28 days in patients with sepsis.

Methods: LOVIT (Lessening Organ dysfunction with VITamin C) is a multicenter, parallel-group, blinded(participants, clinicians, study personnel, Steering Committee members, data analysts), superiority RCT (minimumn = 800). Eligible patients have sepsis as the diagnosis for admission to the intensive care unit (ICU) and arereceiving vasopressors. Those admitted to the ICU for more than 24 h are excluded. Eligible patients arerandomized to high-dose intravenous vitamin C (50 mg/kg every 6 h for 96 h) or placebo. The primary outcome is acomposite of death or persistent organ dysfunction (need for vasopressors, invasive mechanical ventilation, or newand persisting renal replacement therapy) at day 28. Secondary outcomes include persistent organ dysfunction-freedays to day 28, mortality and health-related quality of life at 6 months, biomarkers of dysoxia, inflammation,infection, endothelial function, and adverse effects (hemolysis, acute kidney injury, and hypoglycemia). Six subgroupanalyses are planned.

Discussion: This RCT will provide evidence of the effect of high-dose intravenous vitamin C on patient-importantoutcomes in patients with sepsis.

Trial registration: clinicaltrials.gov, NCT03680274, first posted 21 September 2018.

Keywords: vitamin C, sepsis, septic shock, biomarkers, randomized controlled trial

© The Author(s). 2020 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

* Correspondence: neill.adhikari@utoronto.ca;francois.lamontagne@usherbrooke.ca†Neill K. J. Adhikari and François Lamontagne contributed equally to thiswork.7Department of Critical Care Medicine, Sunnybrook Health Sciences Centre,Toronto, ON, Canada4Centre de Recherche du Centre Hospitalier Universitaire de SherbrookeSherbrooke and Université de Sherbrooke, Sherbrooke, QC, CanadaFull list of author information is available at the end of the article

Masse et al. Trials (2020) 21:42 https://doi.org/10.1186/s13063-019-3834-1

BackgroundThe burden of sepsis is increasing worldwide. Defined as adysregulated host immune response to infection that leadsto organ failure and death [1], sepsis is the cause of up to5.3 million deaths each year globally [2]. Currently, treat-ment options are limited to antimicrobials and organ-supportive care such as intravenous fluids, vasopressors,mechanical ventilation, and renal replacement therapy(RRT). In the absence of effective therapies targeting thedysregulated immune response, prolonged use of life-sustaining therapies is associated with substantial morbidityand mortality [3, 4]. Furthermore, in resource-constrainedsettings, life-sustaining therapies may be unavailable, andthe prognosis of septic patients is grim [5]. This global bur-den led the World Health Assembly to adopt a resolutionurging Member States and the World Health OrganizationDirector-General to take action to reduce the burden ofsepsis through improved prevention, diagnosis, and man-agement [6].Inflammation and oxidative stress are among the main

mechanisms underlying sepsis-induced organ injury anddeath [7]. During sepsis, leukocytes produce reactive oxygenspecies to facilitate killing of pathogens. Normally, endogen-ous antioxidants contain this response and protect cellsfrom oxidative damage. Vitamin C, a key circulating antioxi-dant [8], cannot be synthesized by humans and thereforerequires exogenous intake. Moreover, many critically ill pa-tients are vitamin C deficient and, even when they are not,sepsis further exhausts vitamin C stores [9, 10]. Low serumlevels of ascorbic acid are associated with sepsis-inducedorgan failure and death [11]. Numerous preclinical studieshave shown that, in addition to direct scavenging of oxygenradicals, vitamin C limits their production and restoresendothelial function [12–14]. Vitamin C is also a cofactor inthe synthesis of norepinephrine and vasopressin, hormonesthat are crucial to maintain adequate vascular tone for organperfusion [15].A recent pre-post single-centre study (n = 94) treated

septic shock patients with a combination of intravenous(IV) vitamin C, hydrocortisone, and thiamine and reporteda dramatic reduction in vasopressor requirements, organfailure, and hospital mortality (mortality 8.5% [combinationtherapy] vs. 40.4% [control]; adjusted odds ratio [OR] 0.13,95% confidence interval [CI] 0.04–0.48) [16]. In 2014,Fowler et al. compared high-dose (50mg/kg IV every 6 hfor 4 days) and low-dose vitamin C (12.5mg/kg IV every 6h for 4 days) to placebo in a randomized controlled trial(RCT) of 24 patients with sepsis [17]. High-dose vitamin Cwas associated with a significant improvement in the slopeof Sequential Organ Failure Assessment (SOFA) scores [18]over 96 h, implying faster resolution of organ dysfunction.In 2016, Zabet et al. conducted a 28-patient RCT of vitaminC (25mg/kg IV every 6 h for 72 h) vs. placebo [19] anddemonstrated that vitamin C reduced the mean (± SD)

dose of norepinephrine during the study period (7.44 ±3.65 μg/min vs. 13.79 ± 6.48 μg/min, P = 0.004). A meta-analysis [20] of both trials suggested that intravenousvitamin C monotherapy may reduce mortality (OR 0.21,95%CI 0.04–1.05; P = 0.06), although conclusions werelimited by imprecision and risk of bias (unclear allocationconcealment) of one of the included trials [19]. Morerecently, a RCT of high-dose vitamin C vs. placebo in 167patients with sepsis and acute respiratory distress syndromeshowed no difference in the primary outcomes of change inorgan failure or plasma biomarkers of inflammation andvascular injury [21]. This trial also found an improvementin the secondary outcome of 28-day mortality (risk differ-ence, − 15%, 95%CI − 2% to − 31%, P = 0.03), but the resulthad a very low fragility index of 1 and was one of 46secondary outcomes [22].

Primary aim and hypothesisThe primary aim of the LOVIT (Lessening Organ dysfunc-tion with VITamin C) RCT is to determine whether intra-venous vitamin C, administered as an isolated intervention(not part of a combination of therapies) to patients withsepsis receiving a vasopressor, reduces a composite ofmortality and persistent organ dysfunction [23] (defined asdependency on vasopressors, mechanical ventilation, orincident RRT) at 28 days after randomization, when com-pared to placebo.

MethodsLOVIT is a multicentre, parallel-group, allocation-concealed,blinded (participants, clinicians, study personnel, members ofthe Executive and Steering Committees, and data analysts),superiority RCT. Table 1 shows a timeline of trial activities.The SPIRIT checklist is available in Additional file 2.

Study settingThis trial is centrally coordinated by the Unité deRecherche Clinique et Épidémiologique (URCE) at theCentre de Recherche du Centre Hospitalier Universitairede Sherbrooke, Sherbrooke, QC, Canada. The URCE alsooversees the activities of the LOVIT core laboratory (i.e.storage and analysis of blood and urine samples). TheCLARITY Research Group, Department of Health ResearchMethods, Evidence, and Impact, McMaster University,Hamilton, ON, Canada constructed and will maintain therandomization system and the iDataFax (DF/Net Research,Inc., Seattle, WA, USA) electronic data capture (EDC) sys-tem. The LOVIT trial will be conducted in up to 25 adultintensive care units (ICUs) in Canada and other countries.

Inclusion criteriaPatients will be included if they are:

1) ≥18 years old;

Masse et al. Trials (2020) 21:42 Page 2 of 12

Table 1 LOVIT trial timeline

TIME POINTS Study Period

Days Days Months

Enrolment/Allocation Post-Allocation

1 2 3 4 5–6 7 8–9 10 11–13 14 15–27 28 6 months

ENROLMENT:

Eligibility screen x

Informed consent x

Allocation x

INTERVENTION:

Vitamin C or placebo x x x x

ASSESSMENTS:

Baseline variables

Source of infection x

Severity of illness (APACHE II score) x

Pre-existing comorbidities (Charlsoncomorbidity index, Clinical Frailty Score)

x

Vitamin C levela x

Primary outcome

Death or persistent organ dysfunctionb x

Secondary outcomes

Persistent organ dysfunction-free days inICU, up to day 28b

x x x x x x x x x x x x

Mortality at 6 months x

HRQoL (EQ-5D-5 L) x

Global tissue dysoxia(serum lactate concentration)

x x x

Organ function including renal function(SOFA score)

x x x x x x x x

Inflammation biomarkers (IL-1ß, TNF-α, CRP) x x x

Infection biomarkers (PCT) x x x

Endothelial injury biomarkers (TM, ANG-2) x x x

Occurrence of stage 3 AKIc x x x x x x x x x x x x

Acute hemolysis x x x x x x x x x x x x

Hypoglycemiad x x x x x x x x x

Other variables

Protocol adherencee x x x x

Co-interventionsf x x x x x x x x x x x

AKI acute kidney injury, ANG-2 angiopoietin-2, APACHE II score Acute Physiology and Chronic Health Evaluation II score, CRP C-reactive protein, EQ-5D-5 Lquestionnaire EuroQol-5D (EQ-5D), HRQoL: health-related quality of life, ICU intensive care unit, IL-1ß interleukin-1 beta, PCT procalcitonin, SOFA score SequentialOrgan Failure Assessment score, TM thrombomodulin, TNF-α tumor necrosis factor-alphaaMust be collected before the first dose of investigational productbDependency on mechanical ventilation, RRT, or vasopressorscAs defined by KDIGO (Kidney Disease: Improving Global Outcomes) criteria22dAssessed by core lab (during the time participants receive the 16 doses of the investigational product and the 7 days following the last dose)eReceipt of every planned dose according to schedule until completion of 96-h treatment protocol, samples collected per protocol instructions, vitamin C baselinecollected before the first dose of investigational productfAdministration of mechanical ventilation, RRT, vasopressors, corticosteroids, thiamine, intravenous fluids, blood products, sedatives, and antimicrobials. Daily datauntil ICU discharge or 28 days (whichever comes first)

Masse et al. Trials (2020) 21:42 Page 3 of 12

2) admitted to an ICU with proven or suspectedinfection as the main diagnosis;

3) treated with a continuous intravenous vasopressorinfusion (norepinephrine, epinephrine, vasopressin,dopamine, or phenylephrine) at the time ofeligibility assessment and at randomization.

Exclusion criteriaPatients are excluded for the following reasons:

1) > 24 h since ICU admission;2) known glucose-6-phosphate dehydrogenase (G6PD)

deficiency;3) pregnancy;4) known allergy to vitamin C;5) known kidney stones within the past 1 year;6) received any intravenous vitamin C during current

hospitalization, unless incorporated as part ofparenteral nutrition;

7) expected death or withdrawal of life-sustainingtreatments within 48 h;

8) previously enrolled in this study;9) enrolled in a trial for which co-enrolment is not

possible (to be determined on a case-by-case basisby discussion with the other trial’s principalinvestigators).

Rationale for eligibility criteriaEligibility criteria are broad and consider most patientsadmitted to the ICU with a primary diagnosis of sepsisand receiving vasopressor therapy. We decided not tomandate the Sepsis-3 definition of septic shock [1],which also includes an elevated lactate level, to facilitateenrollment in and enhance generalisability to settingswhere lactate measurements are not routinely available.By including patients admitted to the ICU for ≤24 h, weexclude patients admitted for another reason who de-velop sepsis secondarily and whose prognosis may thusbe more related to the primary reason for ICU admis-sion. By limiting eligibility to patients who are dependenton vasopressors, we are including those with the highestbaseline risk of the primary outcome and therefore thosewith the most to gain from an intervention targeted toincrease synthesis of or sensitivity to endogenous cate-cholamines [15]. The exclusion of patients with knownG6PD deficiency is related to the risks of hemolysis [24];reported cases have occurred with higher doses thanused in LOVIT [25]. Patients with a previous history ofkidney stones may be at increased risk of stone forma-tion with vitamin C due to oxaluria [26].

Recruitment strategy and approach for consentLocal research personnel will screen for potentially eligiblepatients in ICUs and approach eligible patients directly if

they are able to consent. If an eligible patient is notcapable, research personnel will approach the surrogatedecision-maker (SDM) to obtain consent in person, or bytelephone if the SDM is unavailable. Alternatively, thepatient will be randomized and consent will be obtainedsubsequently under a deferred consent model, where per-mitted by the site Research Ethics Board (REB) and as perindividual country requirements concerning informedconsent in medical research. The first dose of the studydrug must be administered within 4 h of randomizationand the blood samples drawn before giving the first doseof study drug.

EthicsThe Coordinating Centre and all participating clinical siteswill receive local REB approval prior to commencing par-ticipant enrolment. Before initiating the trial, each clinicalsite will provide the Coordinating Centre with a copy oftheir local REB approval letter and approved informedconsent form (sample in Additional file 3). Consent is re-quested for future laboratory analyses that may arise fromthis protocol. Any required protocol amendments will besubmitted to each REB and disseminated to all investiga-tors after approval from Health Canada, the regulatoryagency for drug studies in Canada.There is no compensation for harm suffered from trial

participation; details on data collection for adverse events(AEs) are given below. Patients enrolled in this trial arecritically ill and all care will be provided by intensive careclinicians. There is no provision for post-trial care otherthan usual clinical care for ICU patients.

RandomizationA web-based randomization service (www.randomize.net) will be available 24 h per day, 7 days per week. Trialparticipants will be randomized in a 1:1 ratio to vitaminC or matching placebo using permuted blocks of undis-closed and variable size stratified by clinical site.

Trial interventionsThe experimental intervention is intravenous vitamin Cadministered in bolus doses of 50 mg/kg actual bodyweight mixed in a 50-mL solution of either dextrose 5%in water (D5W) or normal saline (0.9% NaCl), givenevery 6 h for 96 h (i.e. 200 mg/kg/day and 16 doses intotal), as long as the patient remains in the ICU. For pa-tients weighing ≥125 kg, 12 ml of D5W or 0.9% NaCl isadded to the bag to keep the concentration of vitamin Cin the range tested in a stability study [27]. For patientsweighing ≥150 kg, the weight is considered as 150 kg tocalculate the dose. Each dose is administered over 30–60min, except for participants > 120 kg, for whom theinfusion time is prolonged so that the rate does notexceed 100 mg/min. Once mixed in either solution,

Masse et al. Trials (2020) 21:42 Page 4 of 12

covered bags may be stored at 4 °C for a maximum of 9days. This procedure was validated by a stability study,which found that at concentrations between 37 and 92mg/mL, vitamin C is physically and chemically stable forat least 14 days at 4 °C, when protected from light [27].Notwithstanding the results of the stability study, the 9-day maximum is derived from regulations concerningintravenous drug administration in Canada [28]. Becausethe infusion time is short, covering medication bags andtubing is not required during the infusion.Participants in the control arm will receive D5W or

0.9% NaCl in a volume to match the vitamin C. Placebowill be infused over the same period as per the instruc-tions for vitamin C and is identical in colour and otherphysical properties to vitamin C.At each clinical site, the preparation of solutions

administered to participants will be the responsibility ofthe unblinded study pharmacist, who is not involved inpatient care. Both vitamin C and placebo (D5W or 0.9%NaCl) will be purchased by sites from their usual dis-tributor. Alternative arrangements, such as shipping ofvitamin C from Canada, may be considered for anyinternational sites.

Rationale for interventionsWe selected a high-dose vitamin C regimen since this maybe associated with the highest likelihood of demonstratingan effect, if one exists [17]. We considered, but rejected, theidea of evaluating the effect of a combined experimentalintervention consisting of vitamin C, corticosteroids, andthiamine. We believe that existing data do not mandate co-administration of vitamin C with other interventions andthat a triple-dummy design would be overly complicatedand may compromise equipoise given recent guidelines infavor of corticosteroid therapy for sepsis [29]. Although thistrial will evaluate vitamin C monotherapy, we will collect in-formation on clinicians’ use of co-interventions includingthiamine and corticosteroids.

Co-interventionsAll co-interventions are at the discretion of the treatingphysicians, who will be blinded to treatment assignment.However, we will record use of relevant co-interventions –ventilation, RRT, vasopressors, corticosteroids, thiamine,nutrition, blood products, analgesia and sedation, and anti-microbials – to detect and report any imbalances betweentreatment groups.Unblinded administration of IV open-label vitamin C,

except as part of parenteral nutrition, is discouraged andwill constitute a protocol violation.

BlindingBlinding will be achieved by using indistinguishablevitamin C and placebo products.

Pharmacists and technicians preparing the study drug(vitamin C or placebo) at each participating site will beunblinded. If a clinical situation calls for unblinding ofother personnel, as assessed on a case-by-case basis, adesignated unblinded Coordinating Centre research co-ordinator will apply the unblinding procedure (Table 2).Unblinded pharmacists will not be authorized to inde-pendently unblind additional personnel.

Primary outcome measureThe primary outcome will be a composite of death orpersistent organ dysfunction (defined as dependency onvasopressors, mechanical ventilation, or new and persist-ing RRT) at day 28 after randomization [23]. Patients withpre-existing end-stage renal disease receiving RRT beforethe index hospitalization will not meet criteria for persist-ent organ dysfunction on the basis of ongoing RRT.

Secondary outcome measuresNote that day 1 refers to the day of randomization.

Table 2 Unblinding of clinical site personnel for emergencymedical management

1. In the event of a medical emergency that directly affects the healthstatus of the participant, it may become necessary to unblind allocationstatus to determine the specific treatment the participant has receivedwhile enrolled in the study. A medical emergency is defined as an eventwhich necessitates immediate attention regarding the treatment of aparticipant.

2. Clinical site personnel (e.g. local principal investigators, co-investigators, and research coordinators) will contact one of the LOVITprincipal investigators if they believe that unblinding of a study partici-pant is medically necessary.

3. The LOVIT principal investigator will discuss the participant’s medicalevent with the clinical site personnel and determine if it is necessary tounblind her/him to the participant’s treatment allocation. At no time willthe participant’s health be compromised or medical treatment delayed.

4. Once approved, the principal investigator or coordinating centreproject leader will contact the designated unblinded researchcoordinator at the coordinating centre, who will provide the clinical sitepersonnel with the participant’s vitamin C or placebo allocation. Thisinformation should be provided by telephone to reduce the risk ofunblinding additional team members.

5. The unblinded research coordinator will complete the LOVITunblinding log. The unblinded research coordinator may contact theresearch coordinator at the clinical site to request any additionalinformation required to complete this log.

6. The unblinded research coordinator is not to unblind the principalinvestigators or any blinded members of the LOVIT team unlessdeemed necessary by the principal investigator. Similarly, clinical sitepersonnel are also not to unblind any other members of the LOVITstudy team (including the Methods Centre research coordinator,principal investigators, or any clinical site research personnel who werenot unblinded) unless deemed necessary by the principal investigator(s).

7. LOVIT personnel must keep all information related to the individualunblinding cases confidential.

8. All cases of unblinding must be documented, including clinical siteID, study ID, date of unblinding, parties unblinded, and reason forunblinding.

Masse et al. Trials (2020) 21:42 Page 5 of 12

Efficacy outcomes will be:

1) Persistent organ dysfunction-free days in ICU,defined as number of days alive and not dependenton vasopressors, mechanical ventilation, or RRT, upto day 28;

2) Mortality at 6 months;3) Health-related quality of life (HRQoL) in 6-month

survivors as assessed by the EuroQol-5D-5 L(EQ-5D-5 L) [30], a standardized questionnaire ofhealth status developed to provide a simple, genericmeasure of health for clinical and economicappraisal. Health status is measured in terms ofmobility, self-care, usual activities, pain/discomfort,and anxiety/depression. Respondents also evaluatetheir overall health status using a 100-point scale;

4) Global tissue dysoxia assessed at days 3 and 7 ofhospitalization by serum lactate concentration [31].This will be assessed using liquid chromatographycoupled with tandem mass spectrometry (LC-MS/MS);

5) Organ function (including renal function) assessedby the SOFA score [18] at days 2, 3, 4, 7, 10, 14,and 28;

6) Inflammation at days 3 and 7 of hospitalization,assessed by serum interleukin-1 beta, tumornecrosis factor-alpha, and C-reactive protein levels,measured by Luminex (Luminex Corp., Austin, TX,USA);

7) An infection biomarker at days 3 and 7 ofhospitalization, measured by serum procalcitonin[17] using an enzyme-linked immunosorbent assay;

8) Endothelial injury assessed at days 3 and 7 ofhospitalization by serum thrombomodulin [17] andangiopoietin-2 [32], measured by Luminex.

Safety outcomes will be:

9) Stage 3 acute kidney injury (AKI) as defined byKidney Disease-Improving Global Outcomes criteria[33] using either creatinine or urine output criteria,at any time during the ICU stay;

10) Acute hemolysis, ascertained until 12 h after the lastdose of study medication, defined as clinicianjudgment of hemolysis, as recorded in the chart, ora hemoglobin drop of at least 25 g/L within 24 h ofa dose of study drug and 2 of the following:reticulocyte count > 2 times upper limit of normal;haptoglobin less than lower limit of normal; indirect(unconjugated) bilirubin > 2 times upper limit ofnormal; or lactate dehydrogenase > 2 times upperlimit of normal. Normal values are as defined ateach participating centre’s laboratory. Severehemolysis is defined as hemoglobin < 75 g/L, at least

2 of the above criteria and the requirement fortransfusion of at least 2 units of packed red bloodcells.

11) Hypoglycemia, defined as a blood glucose levelmeasured in the hospital core laboratory of lessthan 3.8 mmol/L. Vitamin C therapy may beassociated with false hyperglycemic readings whenpoint-of-care glucometers are used to measureblood glucose [34]. Accordingly, blood glucose canonly be measured by one of the following threemethods: hospital core laboratory instruments; apoint-of-care arterial blood gas machine whoseglucose measurement has been validated in thesetting of high blood concentration of ascorbic acid;and point-of-care StatStrip glucometers (NovaBiomedical, Waltham, MA, USA), whosemeasurements have been shown to be accurate inthe presence of high blood concentration ofascorbic acid. Blood glucose measurements inpatients receiving insulin or oral hypoglycemicagents will be monitored as described above up to7 days after the last dose of the investigationalproduct. Because hyperglycemic readings mayprompt iatrogenic hypoglycemic episodes if insulinor oral hypoglycemic agents are administered,hypoglycemic events will be reported as a safetyoutcome.

Adverse eventsIn the context of critical illness, all patients eligible for theLOVIT trial are at risk of AEs. Following Canadian guide-lines for AE reporting in academic critical care trials [35],expected AEs (death, stage 3 AKI, hemolysis, hypoglycemia),whether severe or not, are pre-specified trial outcomes andwill not be reported separately as AEs.In contrast, unexpected AEs that are serious (i.e. fatal,

life-threatening, prolonging hospital stay, resulting inpersistent or significant disability or incapacity, or con-stituting an important medical event according to thelocal principal investigator) and considered by the localprincipal investigator to be at least possibly related totrial procedures will be reported to the CoordinatingCentre within 24 h of becoming aware of the event. TheCoordinating Centre will inform applicable regulatoryagencies of serious AEs possibly related to the trial pro-cedures within 15 days of receiving the report if the AEwas neither fatal nor life-threatening, or within 7 days ifit was fatal or life-threatening.Within 8 days of informing the regulatory agency of a

serious unexpected AE, the Coordinating Centre will sub-mit to the regulatory agency a report including an assess-ment of the importance and implication of findings. Thisinformation will be updated with the patient’s final status.

Masse et al. Trials (2020) 21:42 Page 6 of 12

Additional analysesConditional upon obtaining the participants’ consent forfuture analyses, blood and urine specimens collected inthe context of this trial will be stored for future analyses.These analyses will not include genetic testing.

Data collectionClinical (non-biomarker) data will be obtained from par-ticipants’ medical records. Local research teams will beresponsible for data collection while participants are stillhospitalized. To ensure uniform use of the follow-upquestionnaire and reduce burden on local researchteams, telephone follow-up at 6 months will be con-ducted by the Coordinating Centre research team, or bya designated centre in each country for internationalsites.We will collect the following data:

1) Baseline (day 1): patient demographics, biomarkers(as listed in secondary outcomes), source ofinfection, severity of illness (APACHE II [AcutePhysiology and Chronic Health Evaluation II] score[36]), organ function, including renal function(SOFA score [18]), pre-existing comorbidities asdefined in the Charlson comorbidity index [37], andClinical Frailty Scale estimated within 3 months ofthe current admission [38–40]. Baseline bloodsamples will be collected before the first dose ofstudy product. We will use baseline ascorbic acidlevel for subgroup analysis, described below.

2) Daily data until ICU discharge or 28 days(whichever comes first): protocol adherence(administration of every planned dose untilcompletion of the treatment protocol or ICUdischarge, whichever comes first), co-interventions(ventilation, RRT, vasopressors, corticosteroids,thiamine, nutrition, blood products, analgesia andsedation, and antimicrobials).

3) Days 3 and 7 (if in ICU): blood samples forbiomarkers (as listed in secondary outcomes); urinesamples (for oxalate measurements if requested bythe Data and Safety Monitoring Committee[DSMC]).

4) Days 2, 3, 4, 7, 10, 14, and 28 (if in ICU): SOFAscore [18].

5) 28 days: death or persistent organ dysfunction(defined as dependency on mechanical ventilation,new and persistent RRT, or vasopressors) [23].

6) 6 months: mortality and HRQoL [30]. We willdetermine the date of death for patients that havedied by 6 months.

To minimize participant discomfort and bedsideteams’ workload, study sample collection will coincide as

much as possible with drawing of blood for clinicalreasons.The LOVIT core laboratory will measure the bio-

markers listed as secondary outcomes. Individual sites willship all blood samples to the LOVIT core laboratory foranalysis. Treating teams will remain blinded to the resultsof the biomarker assays performed as part of the study butwill have access to all blood tests obtained for clinical rea-sons. Local laboratory results will be used as required forthe components of the SOFA score. In case of missingdata, SOFA scores will be imputed as pre-specified in thestatistical analysis plan before unblinding the trial.

Study follow-up and cohort retentionOnce a patient is enrolled in the trial, the clinical site willmake every reasonable effort to follow the participant forthe entire duration of the study period. To minimize lossto follow-up at 6months, consent forms will include per-mission to collect alternate contact information. If neces-sary, the Coordinating Centre will request the assistanceof the local research team.Participants may discontinue participation in the LOVIT

trial at any time. If a participant wishes to withdraw con-sent, we will use the following strategies to minimize theimpact on the trial, while respecting autonomy. We willseek a better understanding of the participant’s wishes andoffer the following alternatives to complete withdrawal,which would include no further study drug exposure, datadeletion, and sample destruction:

1) Discontinue study drug but allow data collection(clinical data, sample collection, telephone follow-up);

2) Discontinue study drug, in-person follow-up, andsample collection but allow telephone follow-up;

3) Discontinue study drug, sample collection, and in-person and telephone follow-up, but allow access tomedical records.

Intention to treat and ineligible participantsWe will adhere to the intention-to-treat principle, anddata from participants will be analyzed in the group towhich they will have been allocated irrespective of proto-col adherence. Reasons for protocol deviations, shouldthey arise, will be recorded. If ineligible participants arerandomized, we will allow post-randomization exclusionsif 1) the information about ineligibility was available atrandomization; 2) two members of the Steering Commit-tee blinded to allocation agree that the participant wasmistakenly randomized after review of information frommedical records available at the time of randomization; 3)participants did not receive the assigned intervention; and4) participants remain blinded to their allocation [41, 42].Verification from the Coordinating Center is needed forofficial confirmation of participant withdrawal.

Masse et al. Trials (2020) 21:42 Page 7 of 12

Reducing biasRisk of bias will be reduced by concealed randomizationusing variable and undisclosed blocks. Due to blindingof all relevant parties, decisions to discontinue life-sustaining therapies and other outcomes that requiresubjective assessments (e.g. HRQoL) will not be affectedby individually held beliefs regarding the effects of vita-min C. In addition, to ensure consistent measurement ofbiomarkers in this trial, samples will be collected, frozenon-site and then shipped to the LOVIT core laboratoryin Sherbrooke, QC, Canada, where they will be stored,batched and processed at the end of the trial. Finally, wewill record co-interventions to detect performance bias.

Statistical analysesSample sizeWe determined a minimum sample size of 800 partici-pants based on the following assumptions. We establishedthat an absolute difference of 10% in the composite out-come of death or persistent organ dysfunction (15 to 25%relative risk reduction) is plausible [16, 19] and sufficientlylarge to change practice. Based on recent clinical trials in asimilar population [43], the risk of 28-day persistent organdysfunction or mortality in the control arm is expected tobe approximately 50%. By enrolling 385 evaluable patientsper arm, the study will have 80% power to detect a 10%absolute risk reduction (from 50% to 40%, which corre-sponds to a 20% relative risk reduction). To account forconsent withdrawal and loss to follow-up, we will enroll400 patients per arm. If the control event rate were greateror lower than 50% and the absolute treatment effectremained 10%, statistical power would increase.If international sites join the trial, the sample size and

power to detect a difference in the primary outcomemay increase (see section on International collaborationsand sample size flexibility).

Data analysisAll analyses will be described in detail in a separatelypublished statistical analysis plan. Statistical tests will betwo-sided, and all analyses will be performed using SAS9.4 (SAS Institute, Cary, NC, USA) or later. P < 0.05 willbe interpreted as statistically significant.For the primary analysis, we will compare the propor-

tion of patients meeting the composite endpoint ofpersistent organ dysfunction or death at day 28 betweengroups using a model that accounts for the stratificationvariable [44]. The testing strategy for the primary out-come will maintain an overall two-sided type I error of0.05. We will conduct sensitivity analyses adjusting forimportant pre-specified baseline clinical variables associ-ated with the primary outcome [45]. Subgroup analyseswill use models with terms for subgroup indicators andtheir interaction with treatment.

Pre-specified subgroup analysesThe LOVIT trial will evaluate the effect of vitamin C inthe six subgroups defined at baseline by age (< 65 vs. ≥65years), sex, frailty (Clinical Frailty Scale 1–4 vs. ≥5), sever-ity of illness (quartiles of predicted risk of death frombaseline APACHE II score), strict Sepsis-3 definition ofseptic shock (vasopressor therapy required to maintain amean arterial pressure of 65mmHg and lactate ≥2mmol/L, vs. vasopressor need alone), and baseline ascorbic acidlevel (assuming sufficient variation). We hypothesize thatvitamin C is more beneficial in elderly patients, in thosewith greater frailty and illness severity at baseline, thosewho meet strict criteria for septic shock, and in those withlower baseline ascorbic acid levels.

Interim analysis planThe DSMC will review data on all possibly related ser-ious AEs, hemolysis, stage 3 AKI, and hypoglycemia afterthe enrollment, 28-day follow-up, and data cleaning of250 and 530 patients. If the one-sided P value is < 0.1 (inthe direction of harm in the vitamin C arm) for any ofthe three safety outcomes, an interim two-sided analysisof the primary outcome will be conducted. The DSMCmay also request an analysis of the primary outcome atany time. This analysis will generate a conditional powerfor showing statistically significant efficacy in the finalanalysis of the primary outcome, assuming that thegroup-specific event rates observed to date remain thesame in the total sample size. If the conditional powerfor efficacy is < 20%, in the context of a one-sided P <0.1 for any of the safety outcomes, then the DSMC mayrecommend stopping the trial to the Steering Commit-tee. The DSMC may make a similar recommendationeven if these exact thresholds are not met, based on itsinterpretation of the balance between safety and efficacy.Following the recent RCT of high-dose vitamin C [21],

the DSMC has requested an interim analysis of 28-daymortality after the enrollment of 530 patients, and mayrecommend stopping the trial to the Steering Committeeif two-sided P < 0.001.

RegistrationAn initial ‘no objection’ letter was received from HealthCanada on 5 September 2018 (HC6–24-c219212), andthe trial was registered on www.clinicaltrials.gov on 21September 2018 (NCT03680274).

Data managementThe paper or electronic case report forms (CRFs) are theprimary data collection tool for the study. All data re-quested on the CRF must be recorded either on paperCRFs or on the electronic CRFs within the secure iData-Fax EDC system. If the data are first collected on paperCRFs, site research personnel will subsequently transfer

Masse et al. Trials (2020) 21:42 Page 8 of 12

all data into iDataFax by direct entry. CLARITY researchstaff are responsible for programming and maintainingthe database. With support from CLARITY, Coordinat-ing Centre staff are responsible for trial monitoring.

MonitoringQuality control measures include 1) on-site training ofresearch and clinical personnel; 2) standard operating pro-cedures to guide storage, preparation and administration ofthe study drug as well as processing, storage, and shippingof blood and urine samples; 3) ongoing assessment of trialmanagement metrics (monthly screening logs, weekly re-ports [site enrolment rate, protocol adherence regardingstudy drug administration and study samples, length ofICU stay, vasopressor treatment duration, in-ICU and in-hospital mortality, mortality and persistent organ dysfunc-tion at day 28]) and periodic feedback to the clinical siteson performance (recruitment, protocol adherence), withbenchmarking from other sites; 4) site monitoring visits (re-motely or in person for two of the first five participants and10% of the subsequent participants); 5) ongoing review ofmissing data and outliers; and 6) rapid dissemination of re-sponses to frequently asked questions via our study website(http://lovit.ccctg.ca/) and monthly newsletter. Coordinat-ing Centre staff and the Principal Investigators are availableat all times to answer study-related questions.

Trial oversightExecutive CommitteeThe Executive Committee is comprised of Neill KJ Adhikariand François Lamontagne (co-principal investigators), Marie-Claude Battista (core laboratory), Marie-Hélène Masse, JulieMénard, and Sheila Sprague (co-project leaders). The Execu-tive Committee is responsible for day-to-day managementand is accountable to the Steering Committee.

Steering CommitteeThe Steering Committee consists of Executive Commit-tee members and additional clinical, methodological,and statistical experts, in addition to a patient represen-tative (listed in Additional file 1). It meets quarterly byteleconference. The Steering Committee provides guid-ance and direction to the trial.

Data Safety Monitoring CommitteeThe DSMC is independent of the study investigatorsand responsible for safeguarding the interests of studyparticipants, assessing the safety and efficacy of studyprocedures, and monitoring the overall conduct of thestudy. The DSMC members have extensive trial experi-ence and include a senior methodologist who has servedas Chair on numerous DSMCs for international RCTs, asenior biostatistician, and a clinician-scientist in inten-sive care (Additional file 1). The DSMC will periodically

review enrolment and demographic summaries, protocoldeviations, and serious AEs. In accordance with a pre-specified DSMC Charter, the DSMC will advise the Ex-ecutive and Steering Committees on any concerns relatedto participant safety and trial conduct. After each meeting,the DSMC will make a recommendation for the study tocontinue as designed, termination, continuation withmajor or minor modifications, or temporary suspension ofenrolment until some uncertainty is resolved.

International collaborations and sample size flexibilityLOVIT is designed as a living umbrella RCT, a conceptthat permits the inflation of the overall sample size asinternational collaborations augment trial capacity. Thismodel increases the efficiency of clinical trial research byincreasing the probability that a single trial will be suffi-ciently powered to detect a minimally clinically importanttreatment effect. Because the same core data are collectedin every country, the need to harmonize heterogeneousdatasets is avoided. However, each national lead is free toseek additional funding and to add secondary objectives tothe main protocol.Following this model, the current protocol, CRFs, Op-

eration Manual, and other trial-related documents wereshared with collaborators from other countries who haveagreed to join the LOVIT team of investigators, thuscreating the International LOVIT Collaborative. Ourcollaborators have been invited to become the LOVITnational leads in their respective countries and recruitpatients in the current protocol (n = 800), secure add-itional resources to enroll more patients, or both. Ourterms of engagement stipulate that for participants inthe current protocol (n = 800), specimens for baselinemeasurements of ascorbic acid and lactate must be col-lected and shipped to the LOVIT core laboratory. If ourinternational collaborators secure resources to recruitadditional patients, the clinical data described in thisprotocol will be stored in the Canadian database so thatthe primary analysis may include all participants (n >800). Additional clinical data pertaining to any embed-ded studies may be kept in the country of origin. Theseadd-on studies will focus on outcomes or subgroups notdescribed herein and will be reported separately, coau-thored by the co-principal investigators of this parentprotocol.

DisseminationResults of the LOVIT trial will be presented at inter-national conferences, published in peer-reviewed jour-nals, and disseminated via social media platforms anddiscussion fora managed by partner organizations.The LOVIT protocol is freely accessible via this publi-

cation. The principal investigators, project leaders, andstudy statisticians will have access to the full trial

Masse et al. Trials (2020) 21:42 Page 9 of 12

dataset; there are no contractual limitations to such ac-cess. The LOVIT policy on authorship and sub-studiesthat would require access to the trial dataset is availablein Additional file 4. Other requests for access to the partici-pant-level dataset and statistical code will be consid-ered by the Steering Committee after publication ofprimary results and planned secondary studies by co-investigators.

DiscussionVitamin C, an inexpensive and readily available interven-tion, may be a life-saving therapy in sepsis. If proven ef-fective, vitamin C could be used worldwide to improveoutcomes in high- and low-income settings alike.The LOVIT protocol constitutes a rigorous assessment

of the effect of vitamin C monotherapy on patient-important outcomes in critically ill patients with sepsis.We are working with site principal investigators to con-vince clinicians to enroll eligible patients in LOVIT ra-ther than to prescribe this treatment to patients on thebasis of current data, which constitutes low-certaintyevidence. Once completed, the LOVIT InternationalCollaborative is committed to harmonize data fromLOVIT and other trials of intravenous vitamin C in anindividual patient data meta-analysis.

Trial statusThe current protocol is version 5, last updated 19 September2019. Participant recruitment began on 14 November 2018and is scheduled to continue for 36 months, until ap-proximately 13 November 2021. The database will belocked after the last enrolled patient completes the 6-monthfollow-up, and 6 additional months will be required to ad-dress remaining data queries and to finalize the analyses.Contact information for trial sponsorFrançois Lamontagne (francois.lamontagne@usherbrooke.ca)Université de Sherbrooke3001 12e Avenue NordSherbrooke QC J1H 5 N4 Canada

Supplementary informationSupplementary information accompanies this paper at https://doi.org/10.1186/s13063-019-3834-1.

Additional file 1. LOVIT contributors.

Additional file 2. SPIRIT checklist.

Additional file 3. Model informed consent form.

Additional file 4. Authorship and substudy guidelines.

Abbreviations0.9% NaCI: 0.9% Sodium chloride; AE: Adverse event; AKI: Acute kidney injury;APACHE II: Acute physiology and chronic health evaluation II; CI: Confidenceinterval; CRF: Case report form; D5W: Dextrose 5% in water; DSMC: Data andSafety Monitoring Committee; EDC: Electronic data capture; G6PD: Glucose-6-phosphate dehydrogenase; HRQoL: Health-related quality of life;ICU: Intensive care unit; IV: Intravenous; LOVIT: Lessening Organ dysfunction

with VITamin C; OR: Odds ratio; RCT: Randomized controlled trial;REB: Research Ethics Board; RRT: Renal replacement therapy; SD: Standarddeviation; SDM: Surrogate decision-maker; SOFA: Sequential (sepsis-related)organ failure assessment; URCE: Unité de recherche clinique etépidémiologique

AcknowledgementsWe thank the Unité de Recherche Clinique et Épidémiologique of the Centrede Recherche du Centre Hospitalier Universitaire de Sherbrooke for theircommitment to the coordination of the study; Charles St-Arnaud, MichaelMayette, Joannie Marchand, Marie-Ève Thibault, Brian Beaudoin-Grondin, Marc-André Leclair, Dominique Bérard, Hector Quiroz Martinez and Yannick Poulin fortheir collaboration during the vanguard phase of the study; and Berry Fowler,Heleen Oudemans-van Straaten, Lounes Haroune and Éric Marsault for helpfuldiscussion on various aspects of the protocol. We thank Michelle Kho for hercareful review of an earlier version of this manuscript on behalf of the CanadianCritical Care Trials Group Grants and Manuscripts Committee.LOVIT team members, including research personnel at clinical sites active atthe time of submission of this manuscript, are listed in Additional file 1.

Authors’ contributionsFL and NKJA drafted the protocol for the LOVIT trial and drafted themanuscript; they contributed equally and co-senior authors. MHM, JM, SS,MCB, DJC, GHG, DKH, SK, RP, AD, DC, DA, SM, RP, YA, BKTV, FD, EC, DM, MH,BR, TM, MC, ML, PA, ED,CD, FL, PL, IW, RP, FC, NM, EBC, BB, SW, and JIcontributed to the protocol and revised the manuscript. MHM, JM, SS, MCB,DJC, GHG, DKH, SK, RP, AD, and DC are on the Steering Committee for LOVIT.All authors read and approved the final manuscript.

FundingLOVIT is funded by a grant from the Lotte and John Hecht MemorialFoundation. The funder had no role in the design of the study, ongoingdata collection, planned data analysis and interpretation, or writing of thismanuscript or of the study protocol.

Availability of data and materialsData sharing is not applicable to this article as no datasets were generatedor analysed during the current study.

Ethics approval and consent to participateThis trial protocol has been approved by the Comité d’éthique de larecherche du Centre intégré universitaire de santé et de services sociaux del’Estrie – Centre hospitalier universitaire de Sherbrooke (reference MP-31-2019-2945).Informed consent will be obtained as follows. Local research personnel willapproach eligible patients directly if they are able to consent. If the eligiblepatient is not capable, research personnel will approach the SDM to obtainconsent in person, or by telephone if the SDM is unavailable. Alternatively,the patient will be randomized and consent will be obtained subsequentlyunder a deferred consent model, where permitted by the site REB and asper individual country requirements concerning informed consent inmedical research.All personal health information collected during the study will remain strictlyconfidential in a secure database. Participants will be identified by analphanumeric code, and the linkage from the alphanumeric code toidentifying information will be kept in secure storage under the supervisionof the local principal investigator.

Consent for publicationNot applicable.

Competing interestsThe authors declare that they have no competing interests.

Author details1Centre de Recherche du Centre Hospitalier Universitaire de Sherbrooke,Sherbrooke, QC, Canada. 2Division of Orthopaedic Surgery, Department ofSurgery, McMaster University, Hamilton, ON, Canada. 3Department of HealthResearch Methods, Evidence, and Impact, McMaster University, Hamilton, ON,Canada. 4Centre de Recherche du Centre Hospitalier Universitaire deSherbrooke Sherbrooke and Université de Sherbrooke, Sherbrooke, QC,

Masse et al. Trials (2020) 21:42 Page 10 of 12

Canada. 5Department of Critical Care Medicine, Queen’s University, Kingston,ON, Canada. 6Departments of Pharmacy and Critical Care, The OttawaHospital and Clinical Epidemiology Program, The Ottawa Hospital ResearchInstitute, Ottawa, ON, Canada. 7Department of Critical Care Medicine,Sunnybrook Health Sciences Centre, Toronto, ON, Canada. 8Kingston GeneralHealth Research Institute, Kingston, ON, Canada. 9No institutional affiliation,No institutional affiliation, Sherbrooke, QC, Canada. 10General Intensive CareUnit, Raymond Poincaré Hospital (AP-HP), Lab Inflammation & Infection,U1173 University Paris Saclay-UVSQ/INSERM, Garches, France.11Cardiothoracic and Vascular ICU, Auckland City Hospital, Auckland, NewZealand. 12Medical Research Institute of New Zealand, Wellington, NewZealand. 13School of Nursing, The University of Auckland, Auckland, NewZealand. 14Department of Pathology and Biomedical Science, University ofOtago, Christchurch, New Zealand. 15College of Medicine, King Saud BinAbdulaziz University for Health Sciences, King Abdullah International MedicalResearch Center, Riyadh, Saudi Arabia. 16Department of Critical CareMedicine, Apollo Hospitals, Chennai, India. 17Department of Critical CareMedicine, Queen’s University and Kingston Health Sciences Centre, Kingston,ON, Canada. 18Department of Medicine and Department of Health ResearchMethods, Evidence, and Impact, McMaster University, Hamilton, ON, Canada.19Juravinski Hospital, Hamilton, ON, Canada. 20Centre de Recherche duCentre Hospitalier Universitaire de Montréal, Montréal, QC, Canada. 21Facultéde Médecine, Université Laval, Québec City, QC, Canada. 22Département dessoins intensifs du Centre intégré de santé et des services sociaux deChaudière-Appalaches (Secteur Alphonse-Desjardins), Lévis, QC, Canada.23Department of Medecine, Université Laval and Centre de recherche del’Institut universitaire de cardiologie et de pneumologie de Québec, QuébecCity, QC, Canada. 24Institut Universitaire de Cardiologie et de Pneumologiede Québec, Québec City, QC, Canada. 25Ottawa Hospital Research Institute,Ottawa, ON, Canada. 26Department of Critical Care Research, St. Joseph’sHealthcare, Hamilton, ON, Canada. 27Division of Cardiology, Department ofMedicine, McMaster University, Population Health Research Institute,Hamilton, ON, Canada. 28Centre Hospitalier Universitaire de Sherbrooke,Sherbrooke, QC, Canada. 29Department of Pharmacy, Sunnybrook HealthSciences Centre, Toronto, Canada and Leslie Dan Faculty of Pharmacy,University of Toronto, Toronto, ON, Canada. 30Department of Pharmacy,Sunnybrook Health Sciences Centre, Toronto, ON, Canada.31Interdepartmental Division of Critical Care Medicine and Institute for HealthPolicy, Management, and Evaluation, University of Toronto, Toronto, ON,Canada.

Received: 21 August 2019 Accepted: 22 October 2019

References1. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer

M, Bellomo R, Bernard GR, Chiche JD, Coopersmith CM, et al. The ThirdInternational Consensus Definitions for Sepsis and Septic Shock (Sepsis-3).JAMA. 2016;315(8):801–10.

2. Fleischmann C, Scherag A, Adhikari NK, Hartog CS, Tsaganos T, Schlattmann P,Angus DC, Reinhart K. International Forum of Acute Care Trialists: Assessmentof Global Incidence and Mortality of Hospital-treated Sepsis. Current Estimatesand Limitations. Am J Respir Crit Care Med. 2016;193(3):259–72.

3. Herridge MS, Chu LM, Matte A, Tomlinson G, Chan L, Thomas C, FriedrichJO, Mehta S, Lamontagne F, Levasseur M, et al. The RECOVER Program:disability risk groups and 1-year outcome after 7 or more days ofmechanical ventilation. Am J Respir Crit Care Med. 2016;194(7):831–44.

4. Garland A, Olafson K, Ramsey CD, Yogendran M, Fransoo R. A population-based observational study of intensive care unit-related outcomes. Withemphasis on post-hospital outcomes. Ann Am Thorac Soc. 2015;12(2):202–8.

5. Dugani S, Veillard J, Kissoon N. Reducing the global burden of sepsis. CMAJ.2017;189(1):E2–3.

6. Reinhart K, Daniels R, Kissoon N, Machado FR, Schachter RD, Finfer S.Recognizing sepsis as a global health priority - a WHO resolution. N Engl JMed. 2017;377(5):414–7.

7. Angus DC, van der Poll T. Severe sepsis and septic shock. N Engl J Med.2013;369(9):840–51.

8. Frei B, Stocker R, England L, Ames BN. Ascorbate: the most effectiveantioxidant in human blood plasma. Adv Exp Med Biol. 1990;264:155–63.

9. Amrein K, Oudemans-van Straaten HM, Berger MM. Vitamin therapy incritically ill patients: focus on thiamine, vitamin C, and vitamin D. IntensiveCare Med. 2018;44(11):1940–4.

10. Carr AC, Rosengrave PC, Bayer S, Chambers S, Mehrtens J, Shaw GM.Hypovitaminosis C and vitamin C deficiency in critically ill patients despiterecommended enteral and parenteral intakes. Crit Care. 2017;21(1):300.

11. de Grooth H, Spoelstra-de Man A, Oudemans-van Straaten H. Early plasmavitamin C concentration, organ dysfunction and ICU mortality [abstract0723]. Intensive Care Med. 2014;40:S199.

12. May JM, Harrison FE. Role of vitamin C in the function of the vascularendothelium. Antioxid Redox Signal. 2013;19(17):2068–83.

13. Oudemans-van Straaten HM, Spoelstra-de Man AM, de Waard MC. VitaminC revisited. Crit Care. 2014;18(4):460.

14. Wilson JX. Evaluation of vitamin C for adjuvant sepsis therapy. AntioxidRedox Signal. 2013;19(17):2129–40.

15. Carr AC, Shaw GM, Fowler AA, Natarajan R. Ascorbate-dependentvasopressor synthesis: a rationale for vitamin C administration in severesepsis and septic shock? Crit Care. 2015;19:418.

16. Marik PE, Khangoora V, Rivera R, Hooper MH, Catravas J. Hydrocortisone,vitamin C, and thiamine for the treatment of severe sepsis and septic shock:a retrospective before-after study. Chest. 2017;151(6):1229–38.

17. Fowler AA 3rd, Syed AA, Knowlson S, Sculthorpe R, Farthing D, DeWildeC, Farthing CA, Larus TL, Martin E, Brophy DF, et al. Phase I safety trialof intravenous ascorbic acid in patients with severe sepsis. J TranslMed. 2014;12:32.

18. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H,Reinhart CK, Suter PM, Thijs LG. The SOFA (Sepsis-related Organ FailureAssessment) score to describe organ dysfunction/failure. On behalf of theWorking Group on Sepsis-Related Problems of the European Society ofIntensive Care Medicine. Intensive Care Med. 1996;22(7):707–10.

19. Zabet MH, Mohammadi M, Ramezani M, Khalili H. Effect of high-doseascorbic acid on vasopressor's requirement in septic shock. J Res PharmPract. 2016;5(2):94–100.

20. Langlois PL, Manzanares W, Adhikari NKJ, Lamontagne F, Stoppe C, Hill A,Heyland DK. Vitamin C administration to the critically ill: a systematic reviewand meta-analysis. JPEN J Parenter Enteral Nutr. 2019;43(3):335–46.

21. Fowler AA 3rd, Truwit JD, Hite RD, Morris PE, DeWilde C, Priday A, Fisher B,Thacker LR 2nd, Natarajan R, Brophy DF, et al. Effect of Vitamin C infusionon organ failure and biomarkers of inflammation and vascular injury inpatients with sepsis and severe acute respiratory failure: The CITRIS-ALIRandomized Clinical Trial. JAMA. 2019;322(13):1261–70.

22. Walsh M, Srinathan SK, McAuley DF, Mrkobrada M, Levine O, Ribic C, MolnarAO, Dattani ND, Burke A, Guyatt G, et al. The statistical significance ofrandomized controlled trial results is frequently fragile: a case for a FragilityIndex. J Clin Epidemiol. 2014;67(6):622–8.

23. Heyland DK, Muscedere J, Drover J, Jiang X, Day AG. Canadian Critical CareTrials G: Persistent organ dysfunction plus death: a novel, compositeoutcome measure for critical care trials. Crit Care. 2011;15(2):R98.

24. Cappellini MD, Fiorelli G. Glucose-6-phosphate dehydrogenase deficiency.Lancet. 2008;371(9606):64–74.

25. Campbell GD Jr, Steinberg MH, Bower JD. Letter: Ascorbic acid-inducedhemolysis in G-6-PD deficiency. Ann Intern Med. 1975;82(6):810.

26. de Grooth HJ, Manubulu-Choo WP, Zandvliet AS, Spoelstra-de Man AME,Girbes AR, Swart EL, Oudemans-van Straaten HM. Vitamin Cpharmacokinetics in critically ill patients: a randomized trial of four IVregimens. Chest. 2018;153(6):1368–77.

27. Walker SE, Iazzetta J, Law S, Kanji S, Bolduc B, Lamontagne F, Adhikari NKJ.Administration of intravenous ascorbic acid-practical considerations forclinicians. Nutrients. 2019;11(9):1994.

28. National Association of Pharmacy Regulatory Authorities. Model standardsfor pharmacy compounding of non-hazardous sterile preparations (revisionNovember 2016). Ottawa, Canada: National Association of PharmacyRegulatory Authorities; 2015.

29. Lamontagne F, Rochwerg B, Lytvyn L, Guyatt GH, Moller MH, Annane D,Kho ME, Adhikari NKJ, Machado F, Vandvik PO, et al. Corticosteroid therapyfor sepsis: a clinical practice guideline. BMJ. 2018;362:k3284.

30. Herdman M, Gudex C, Lloyd A, Janssen M, Kind P, Parkin D, Bonsel G, BadiaX. Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Qual Life Res. 2011;20(10):1727–36.

Masse et al. Trials (2020) 21:42 Page 11 of 12

31. Gu WJ, Zhang Z, Bakker J. Early lactate clearance-guided therapy in patientswith sepsis: a meta-analysis with trial sequential analysis of randomizedcontrolled trials. Intensive Care Med. 2015;41(10):1862–3.

32. Yeo TW, Lampah DA, Gitawati R, Tjitra E, Kenangalem E, Piera K, Price RN,Duffull SB, Celermajer DS, Anstey NM. Angiopoietin-2 is associated withdecreased endothelial nitric oxide and poor clinical outcome in severefalciparum malaria. Proc Natl Acad Sci U S A. 2008;105(44):17097–102.

33. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney InjuryWork Group. KDIGO Clinical Practice Guideline for Acute Kidney Injury.Kidney Int Suppl. 2012;2:1–138.

34. Kahn SA, Lentz CW. Fictitious hyperglycemia: point-of-care glucosemeasurement is inaccurate during high-dose vitamin C infusion for burnshock resuscitation. J Burn Care Res. 2015;36(2):e67–71.

35. Cook D, Lauzier F, Rocha MG, Sayles MJ, Finfer S. Serious adverse events inacademic critical care research. CMAJ. 2008;178(9):1181–4.

36. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity ofdisease classification system. Crit Care Med. 1985;13(10):818–29.

37. Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combinedcomorbidity index. J Clin Epidemiol. 1994;47(11):1245–51.

38. Rockwood K, Song X, MacKnight C, Bergman H, Hogan DB, McDowell I,Mitnitski A. A global clinical measure of fitness and frailty in elderly people.CMAJ. 2005;173(5):489–95.

39. Bagshaw SM, Stelfox HT, McDermid RC, Rolfson DB, Tsuyuki RT, Baig N,Artiuch B, Ibrahim Q, Stollery DE, Rokosh E, et al. Association between frailtyand short- and long-term outcomes among critically ill patients: amulticentre prospective cohort study. CMAJ. 2014;186(2):E95–102.

40. Muscedere J, Waters B, Varambally A, Bagshaw SM, Boyd JG, Maslove D,Sibley S, Rockwood K. The impact of frailty on intensive care unitoutcomes: a systematic review and meta-analysis. Intensive Care Med.2017;43(8):1105–22.

41. Akl EA, Briel M, You JJ, Sun X, Johnston BC, Busse JW, Mulla S, LamontagneF, Bassler D, Vera C, et al. Potential impact on estimated treatment effects ofinformation lost to follow-up in randomised controlled trials (LOST-IT):systematic review. BMJ. 2012;344:e2809.

42. Fergusson D, Aaron SD, Guyatt G, Hebert P. Post-randomisation exclusions:the intention to treat principle and excluding patients from analysis. BMJ.2002;325(7365):652–4.

43. Heyland D, Muscedere J, Wischmeyer PE, Cook D, Jones G, Albert M, Elke G,Berger MM, Day AG. Canadian Critical Care Trials Group: A randomized trialof glutamine and antioxidants in critically ill patients. N Engl J Med. 2013;368(16):1489–97.

44. Senn SJ, Lewis RJ. Treatment effects in multicenter randomized clinical trials.JAMA. 2019;321(12):1211–2.

45. Committee for Proprietary Medicinal Products (CPMP). Committee forProprietary Medicinal Products (CPMP): points to consider on adjustment forbaseline covariates. Stat Med. 2004;23(5):701–9.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Masse et al. Trials (2020) 21:42 Page 12 of 12